You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
After attempting to delete old Cronicle job logs following the instructions in post #304, the Event History tab in the Cronicle web UI displayed unexpected behavior.
Cleanup Process Used:
/opt/cronicle/bin/control.sh stop
find /opt/cronicle/data/logs/ -type f -name "*.json" -mtime +30 -exec ls -lh {} +
or
find /opt/cronicle/data/logs/ -type d -mtime +30 -exec rm -rf {} +
/opt/cronicle/bin/control.sh start
The cleanup successfully removed logs, but an issue appeared in the Event History:
First tabs displayed jobs normally, but when changing to certain tab, it showed "No events found." as is there no logs for that job
The UI still reports 2,000 jobs, but given that this task runs daily, there should be only around 30 jobs instead.
The last job on the working tab was from January 18, meaning that if logs were deleted correctly, there should be job entries at least until January 11 (30 days before February 11, when the cleanup was performed).
So I have this questions:
Is this method of deleting old logs incorrect?
Is there a database conflict where job metadata still exists but logs are missing?
Or could this be a bug in Cronicle?
Would appreciate clarification on whether this issue is expected or if a different cleanup approach is recommended.
Operating System
Ubuntu 22.04.5 LTS
Node.js Version
v12.22.9
Cronicle Version
0.9.23
Server Setup
Single Server
Storage Setup
Local Filesystem
Relevant log output
Code of Conduct
I agree to follow this project's Code of Conduct
The text was updated successfully, but these errors were encountered:
Oh dear. Yeah, so, those commands from #304 are extremely brute force, and only used as a last resort. It's akin to going into a SQL database and deleting random index files. It's incredibly dangerous and not recommended, except in very rare cases where there is no other option.
Cronicle uses the filesystem as a database (sort of like git does), so there are many different JSON records that act as indexes, tables, etc. Also, it caches them in memory, so if you make any FS changes while Cronicle is running, it will become horribly corrupted.
At this point I have no idea what state the Cronicle FS "database" is in, so I'd recommend you do an export, total wipe of the data dir, then re-import. The export contains everything except job history. Instructions here:
Is there an existing issue for this?
What happened?
After attempting to delete old Cronicle job logs following the instructions in post #304, the Event History tab in the Cronicle web UI displayed unexpected behavior.
Cleanup Process Used:
The cleanup successfully removed logs, but an issue appeared in the Event History:
So I have this questions:
Would appreciate clarification on whether this issue is expected or if a different cleanup approach is recommended.
Operating System
Ubuntu 22.04.5 LTS
Node.js Version
v12.22.9
Cronicle Version
0.9.23
Server Setup
Single Server
Storage Setup
Local Filesystem
Relevant log output
Code of Conduct
The text was updated successfully, but these errors were encountered: